We screened the small molecule compounds that stimulate osteogenesis by themselves or promote bone morphogenetic protein (BMP)-induced bone formation. We found that a specific inhibitor for MAPK/extracellular signal-regulated kinase kinase (MEK)-1, promoted the early osteoblastic differentiation and mineralization of extracellular matrix (ECM) in C2C12 pluripotent mesenchymal cells treated with recombinant human BMP-2 (rhBMP-2) and MC3T3-E1 preosteoblastic cells. ALP activity was synergistically increased by the treatment with a specific MEK-1 inhibitor PD98059 and rhBMP-2 in both cell lines. Twenty-five micromolar PD98059 promoted mineralization of ECM in rhBMP-2-treated C2C12 cells and MC3T3-E1 cells. In contrast, PD98059 reduced osteocalcin (OCN) secretion and its transcriptional level in rhBMP-2-treated C2C12 cells but increased its secretion and mRNA level in MC3T3-E1 cells. Stable expression of a dominant-negative MEK-1 mutant in C2C12 cells represented high ALP activity and low osteocalcin production in the presence of rhBMP-2, while a constitutively active mutant of MEK-1 attenuated both of them. Together, our results indicated that BMP-2-induced mineralization of ECM of pluripotent mesenchymal stem cells and preosteoblastic cells could be controlled by a fine tuning of the MAPK signaling pathway. Further, MEK-1 inhibitors would be useful for the promotion of bone formation, for instance, the treatments for delayed fracture healing or advance of localized osteoporotic change after fracture healing.
Orthopedic surgeons often have opportunities to manage the fractures of osteoporotic bones, especially the distal radius and proximal femur of elderly women. Several complications may occur during the treatment, for instance, the nonunion of fracture site, advance of localized osteoporotic change around the fracture site because of the prolonged period of treatment, etc. To overcome these problems, we screened the small molecule compounds that induced osteogenesis by themselves or promoted bone morphogenetic protein (BMP)-induced bone formation. The final goal of our study would be clinical application of them. Here, we found that a specific inhibitor for MAPK/extracellular signal-regulated kinase kinase (MEK)-1 promoted mineralization of the extracellular matrix (ECM) formed by mesenchymal pluripotent cells and preosteoblastic cells in vitro.
BMPs are members of the transforming growth factor (TGF)-β superfamily and play critical roles in osteogenesis. They induce ectopic bone formation in vivo when implanted in muscular tissue(1) and modulate the osteoblastic differentiation of mesenchymal stem cells in vitro.(2–5) The molecular mechanisms of the osteoblastic differentiation by BMP-2 were well characterized.(6–10) BMP-2 is thought to exert their biological function by interacting with two types of transmembrane serine/threonine kinase receptors. These BMP receptors phosphorylate transcriptional factors Smad1, -5, and -8. The phosphorylated Smads bind to Smad4 following the complex translocates into the nucleus and regulates the transcriptional activation of genes related to the osteoblastic differentiation resulting in bone formation. A clinical trial of recombinant human BMP-2 (rhBMP-2) has been done, but a huge amount of rhBMP-2 was required for bone formation to the treatment of fracture, bone defect, spinal fusion, etc.(11,12) Thus, small molecule compounds, which synergistically reinforce the effects of rhBMP-2 on bone formation or induce osteogenesis by themselves, would be useful for its clinical application. Simvastatin, a 3-hydroxy-3-methylglutaryl coenzyme A (HMG CoA) reductase inhibitor, was first reported to be one candidate that promoted osteogenesis.(13,14)
The MEK/MAPK signaling pathway plays significant roles in cell proliferation and differentiation. This signaling also affects the osteoblastic differentiation in mesenchymal cells. Takeuchi et al. showed that integrin activation by ECM induced the activation of MAPK, which is necessary for the osteoblastic differentiation of MC3T3-E1 preosteoblasts.(15) Jaiswal et al. showed that the osteoblastic differentiation by glucocorticoid dexamethasone of adult human mesenchymal stem cells was inhibited by MEK inhibition.(16) In contrast, Kertzschmar suggested that the activation of MAPK inhibited BMP signaling by phosphorylating the linker region between the Mad homology (MH) 1 domain and MH2 domain of Smad1 and inhibiting nuclear translocation of this protein.(17) In addition, it has been reported that MAPK activation down-regulated type I collagen gene expression in MC3T3-E1 cells.(18) Therefore, the involvement of the MAPK signaling pathway on osteoblastic differentiation is somewhat controversial.
Here, we reported the effects of long-lasting alteration in MAPK signaling on the osteoblastic differentiation of mesenchymal cell lines using a specific MEK-1 inhibitor and stable expression of constitutively active or dominant-negative MEK-1 mutant as assessed by ALP activity, osteocalcin (OCN) secretion, their transcriptional levels, and a mineralized nodule.
MATERIALS AND METHODS
C2C12 pluripotent mesenchymal cells and MC3T3-E1 preosteoblastic cells were purchased from Riken Cell Bank (Tsukuba, Japan). C2C12 cells were cultured in DMEM supplemented with 10% FBS (Equitech-bio, Kerrville, TX, USA) at 37°C in a humidified atmosphere of 5% CO2 and MC3T3-E1 cells in α-minimum essential medium (α-MEM) containing 10% FBS. All mediums were purchased from Invitrogen Life Technologies (Tokyo, Japan).
For each assay, C2C12 cells and stable transfectants were seeded at 1 × 104 cells/cm2, and MC3T3-E1 cells were seeded at 2 × 104 cells/cm2. Twenty-four hours after plating, the medium was replaced by the new medium containing 10% FBS in the absence or presence of rhBMP-2 (generous gift from the Genetics Institute, Cambridge, MA, USA, and Yamanouchi pharmaceutical Co., Tokyo, Japan) and MEK-1 inhibitor PD98059 (New England Biolabs, Inc., Beverly, MA, USA).
Constructs and transfection
Hemagglutinin (HA)-tagged constitutively active rat MEK-1 expression vector was purchased from Upstate Biotechnology (Lake Placid, NY, USA). HA-tagged constitutively active MEK-1 cDNA was excised from the vector as a BamHI/XhoI fragment and transferred into pcDNA3 expression vector (Invitrogen Life Technologies). HA-tagged nonactivatable (dominant-negative) MEK-1 cDNA was made by alanine substitutions of two aspartic acids in HA-tagged constitutively active MEK-1 cDNA, which were substituted in place of serine residues 218 and 222 in wild-type cDNA for its activation, using polymerase chain reaction (PCR) using appropriate primers.(19)
These plasmids and empty vectors were transfected into C2C12 cells using LipofectAMINE PLUS (Invitrogen Life Technologies) following the manufacturer's protocol and selected with 1 mg/ml of G418 sulfate (Invitrogen Life Technologies) to obtain stable transfectants. Expression of mutated protein in stable transfectants was detected by immunoblotting using anti-MEK-1 antibody and anti-HA antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA).
ALP staining and activity
C2C12 cells and MC3T3-E1 cells were treated with or without rhBMP-2 and MEK-1 inhibitor 24 h after being seeded and incubated for 3 days.
For ALP staining, cells were fixed for 15 minutes with 3.7% formaldehyde at room temperature after being washed with PBS. After the fixation, they were incubated with the mixture of nitro blue tetrazolium (NBT; Promega, Madison, WI, USA) and 5-bromo-4-chloro-3-indolyl-phosphate (BCIP; Promega) in ALP buffer (100 mM of Tris-HCl, pH 9.5, 100 mM of NaCl, and 5 mM of MgCl2) for 1 h in the dark at room temperature.
To measure ALP activity, cells were washed twice with PBS and lysed in M-Per Mammalian Protein Extraction Reagent (Pierce, Rockford, IL, USA) following its protocol. ALP activity was assayed using p-nitrophenylphosphate as a substrate by Alkaline Phospha Test Wako (Wako Pure Chemicals Industries, Ltd., Osaka, Japan) and the protein content was measured using the bicinchoninic acid (BCA) protein assay kit (Pierce).
Transfectants were treated with or without rhBMP-2 for 3 days and ALP activity was measured as mentioned previously.
Immunofluorescence for troponin T
C2C12 cells were cultured in DMEM supplemented with 10% FBS in the absence or presence of 300 ng/ml of rhBMP-2 and 25 μM of PD98059 on a Lab-Tek chamber slide (Nunc, Roskilde, Denmark) for 3 days. Cells were fixed with 3.7% formaldehyde for 15 minutes at room temperature and permeabilized with 0.5% Triton-X 100 in PBS. After washing the cells with PBS twice and blocking with 5% normal goat serum in 0.1% Tween 20-containing PBS (T-PBS), they were incubated for 1 h with mouse anti-troponin T (TnT) monoclonal antibody (Sigma, Tokyo, Japan) at 1:200 dilution at room temperature. The cells were washed three times with T-PBS and incubated for 30 minutes with fluorescein isothiocyanate (FITC)-conjugated anti-mouse secondary antibody at 1:200 dilution (Santa Cruz Biotechnology, Inc.). Immunofluorescence for TnT was followed by ALP staining. The cells were observed using a fluorescence microscopy IX-70 (Olympus, Tokyo, Japan) attached with a camera. The number of total TnT+ or ALP+ cells was counted in five microscopic fields of one experiment and the percentage of positive cells/total cells was calculated. Total number of the cells was counted over 2000 in one experiment. Three duplicated experiments were done independently.
The amount of OCN secreted into the culture medium between day 4 and 6 was determined by radioimmunoassay using the mouse osteocalcin immunoradiometric assay (IRMA) kit (Immutopics, Inc., San Clemente, CA, USA). Because the MEK inhibitor showed cellular growth inhibition, OCN secretion was normalized to the total cellular protein content.
Alizarin red S staining and calcium content in mineralized nodule assay
C2C12 cells were cultured in DMEM containing 10% FBS, 50 μg/ml of ascorbic acid (Invitrogen Life Technologies), and 10 mM of β-glycerophosphate (Sigma) in the absence or presence of 300 ng/ml of rhBMP-2 and 25 μM of MEK inhibitor for 12 days for mineralized nodule assay. MC3T3-E1 cells were cultured in α-MEM containing 10% FBS and 10 mM of β-glycerophosphate in the absence or presence of 50 ng/ml of rhBMP-2 and 25 μM of PD98059 for 15 days. The medium was replaced every 3 days. For alizarin red S (Sigma) staining to detect mineralized nodules, the cells were washed with deionized water after being fixed for 15 minutes with 3.7% formaldehyde at room temperature and stained with alizarin red S at pH 6.3.
To measure calcium content of the nodule, 500 μl of 0.6N HCl was added to each well to decalcify mineralized nodules after fixation of the cells.(20) After 24 h, calcium content in the supernatant was determined using the o-cresolphthalein complexon color development method by the Calcium Test Wako (Wako Pure Chemicals Industries, Ltd.). Duplicated wells were used to determine the protein content.
After cells were cultured with various treatments for 3 days, they were lysed rapidly in ice with Laemmli's SDS sample buffer (62.5 mM of Tris-HCl, pH 6.8, 2% SDS, 10% glycerol, and 0.01% phenol red)(21) containing 40 mM of dithiothreitol with a cell scraper. These samples were subjected to SDS 4–20% gradient PAGE (Daiichi Pure Chemicals Co, Ltd., Tokyo, Japan) for detection of MAPK or 10% for TnT, myosin heavy chain (MHC), MyoD, and β-actin. Proteins were transferred to nitrocellulose membranes (Bio-Rad Laboratories, Inc., Hercules, CA, USA). The membranes were immunoblotted with primary antibodies in T-PBS containing 1% bovine serum albumin (BSA; Promega) overnight after 1-h blocking with T-PBS containing 3% BSA. Anti-MAPK antibody (New England Biolabs, Inc.) and anti-phosphoMAPK antibody (New England Biolabs, Inc.) were used as the primary antibody at 1:1000 dilution. Anti-TnT antibody, anti-MHC antibody (MF-20; Developmental Studies Hybridoma Bank, Iowa City, IA, USA), and anti-β-actin antibody (Chemicon International, Temecala, CA, USA) were used at 1:200. The membrane was incubated for 30 minutes with secondary antibodies (anti-rabbit or anti-mouse immunoglobulin G [IgG] Fc ALP conjugate; Promega) at 1:7500 dilution after wash for 10 minutes in T-PBS three times. Immunoreactive bands were visualized by incubation of the membrane in the mixture of NBT and BCIP in ALP buffer mentioned previously.
Phosphorylation level of MAPK
For relative estimation of phosphorylation of MAPK, the blot membrane was scanned with a GT9500 flat scanner (Epson, Tokyo, Japan) and analyzed with the National Institutes of Health (NIH) image software.
An immunoreactive band of phosphorylated p42 MAPK was distinguished from that of nonphosphorylated p42 MAPK by SDS 4–20% gradient PAGE. The standard absolute phosphorylation level of p42 MAPK was estimated using the membrane that was immunoblotted by anti-MAPK antibody as the following equation:
The signal closest to 50% was used to calculate a standard absolute phosphorylation level of MAPK and avoided the interference with pp42 and p42 immunoreactive band signal when the large difference exists between these two.(22) Relative phosphorylation level was estimated from the signal of phosphorylated p42 in membrane blotted by anti-phospho-MAPK antibody, which was normalized with the signal of total p42. Each absolute phosphorylation level was estimated as the following equation:
Total RNA was extracted using TRIZOL reagent (Invitrogen Life Technologies). Ten micrograms of total RNA was electrophoresed in 1% agarose-formaldehyde gels and transferred onto Hybond N+ nylon membrane (Amersham Pharmacia Biotech, Tokyo, Japan). Twenty-five nanograms of the probes was radiolabeled with [α-32P]deoxycytidine 5′-triphosphate (dCTP) using the Rediprime II DNA Labeling System (Amersham Pharmacia Biotech). The fragments of rat ALP cDNA(23) and mouse OCN cDNA(24) were used as probes. The membranes were prehybridized, hybridized using Rapid-Hyb Buffer (Amersham Pharmacia Biotech) with radioactive probes, and then washed with 2× SSC containing 0.1% SDS or 0.2× SSC containing 0.1% SDS. The hybridized blots were exposed to an imaging plate at room temperature and relative levels of mRNA were calculated by a laser scanning densitometer (Fuji BAS 2000; Fuji, Tokyo, Japan).
Data are expressed as mean ± SD in all figures. Statistical significance was analyzed by Student's t-test.
Effects of the MEK-1 inhibitor on the phosphorylation of MAPK and morphology in C2C12 pluripotent mesenchymal cells
The phosphorylation level of MAPK was reduced by the treatment with MEK-1 inhibitor PD98059 in C2C12 cells in a dose-dependent fashion (60.5 ± 8.26% to 28.9 ± 16.4% at 0–100 μM; Fig. 1A).
Morphologically, C2C12 cells revealed a spindle-shape (Fig. 1B, a and b) and expressed small amount of TnT, a myogenic marker (Fig. 1B, e) cultured in DMEM supplemented with 10% FBS in the absence of rhBMP-2. When cells were treated with 25 μM of MEK-1 inhibitor in the absence of rhBMP-2, the percentage of TnT+ cells was increased ∼4.9-fold (0.87 ± 0.97% to 4.30 ± 1.88%; Fig. 1B, e and f, lower right graph). In the presence of 300 ng/ml of rhBMP-2, the cell shape changed to polygonal (Fig. 1B, c and d). ALP+ cells appeared, whereas TnT+ ones disappeared (Fig. 1B, c, d, g, and h). The percentage of ALP+ cells was increased 1.9-fold after the treatment with 25 μM of PD98059 (11.8 ± 3.34% to 22.5 ± 6.46%; Fig. 1B, c and d, lower left graph). Myogenic differentiation was also detected by the immunoblotting with TnT, MHC, and MyoD antibodies (Fig. 1C). Expression of these myogenic markers was increasing by the treatment with MEK-1 inhibitor even in the presence of 10% FBS. Together, these results indicated that the MEK-1 inhibitor PD98059 stimulated both myogenic and osteogenic differentiation of C2C12 cells.
Effects of the MEK-1 inhibitor on the rhBMP-2-induced osteoblastic differentiation in C2C12 cells
To further evaluate the effects of MEK-1 inhibitor on the osteoblastic differentiation, ALP staining was carried out to visualize the total ALP activity of C2C12 cells (Fig. 2A). rhBMP-2 stimulated ALP activity in C2C12 cells in a dose-dependent fashion (0–300 ng/ml) and ALP activity was increasing also by the treatment with MEK-1 inhibitor in a concentration-dependent manner (0–100 μM). Figure 2B clearly shows that rhBMP-2 and PD98059 synergistically increased ALP activity. RNA blot analysis confirmed the increase of the ALP transcriptional level by the treatment with MEK-1 inhibitor after incubation for both 3 days and 6 days (Fig. 2D, left panel).
Three hundred nanograms per milliliter of rhBMP-2 induced OCN secretion from C2C12 cells. In contrast to ALP activity, OCN secretion from C2C12 cells between day 4 and 6 decreased in the presence of PD98059 in a dose-dependent fashion (0–100 μM; Fig. 2C, right five bars). OCN mRNA level also was decreased slightly by the treatment with PD98059 for 6 days, which was consistent with OCN secretion, and it was increased at day 3 (Fig. 2D, right panel).
Alizarin red S staining indicated that 300 ng/ml of rhBMP-2 induced mineralization of ECM of C2C12 cells and that the additional treatment with 25 μM of MEK-1 inhibitor enhanced the calcium deposition (Fig. 2E, panel). Calcium content in the mineralized nodules formed by these cells stimulated with both 300 ng/ml of rhBMP-2 and 25 μM of PD98059 was increased 1.8-fold as compared with that stimulated with 300 ng/ml of rhBMP-2 only (Fig. 2E).
Effects of overexpression of constitutively active or dominant-negative MEK-1 on rhBMP-2-induced osteoblastic differentiation of C2C12 cells
To further confirm the involvement of MAPK signaling on the osteoblastic differentiation, we established two C2C12 cell clones stably expressing constitutively active MEK-1 (CA1 and CA2), two clones expressing dominant-negative mutant (DN1 and DN2), and one mock vector-expressing clone as a control. Immunoblot analysis confirmed the expression of HA-tagged MEK-1 in CA1, CA2, DN1, and DN2 (Fig. 3A, upper panel). When parental C2C12 cells and five transfectants were cultured in DMEM supplemented with 10% FBS for 3 days, the phosphorylation level of MAPK was 56.4% in mock transfectant and >80% in CA transfectants (83.8% in CA1 and 86.1% in CA2) and relatively lower (42.7% in DN1 and 52.5% in DN2) in DN transfectants (Fig. 3A, middle and lower panels).
Morphologically, DN transfectants showed a fibroblast-like shape and their nucleus shape was clear (Fig. 3B, c and d), and two CA transfectants represented a round to spindle-shape with an unclear nucleus under phase-contrast microscopy (Fig. 3B, e and f).
DN transfectants showed higher ALP activity than the mock transfectant, whether they were treated with or without rhBMP-2 (Fig. 3C). In contrast, both CA1 and -2 transfectants represented little ALP activity even in the presence of 300 ng/ml of rhBMP-2 (Fig. 3C).
Both mock and DN transfectants produced OCN into the culture media by stimulation with 300 ng/ml of rhBMP-2; OCN secretion from DN transfectants was decreased 33% in DN1 and 21% in DN2, as much as that of mock (Fig. 3D). CA transfectants did not secrete detectable OCN into the culture media (data not shown). These results were consistent with those using the MEK-1 inhibitor (Fig. 2).
Effects of the MEK-1 inhibitor on the osteoblastic differentiation of MC3T3-E1 preosteoblastic cells
Next, we examined the effects of the MEK-1 inhibitor on MC3T3-E1 preosteoblastic cells.
The phosphorylation of MAPK was also reduced by PD98059 in MC3T3-E1 preosteoblastic cells in its dose-dependent fashion (79.5 ± 11.6% to 54.8 ± 6.99% at 0 to 100 μM; Fig. 4A).
ALP activity was increased by the treatment with PD98059 in the absence or presence of 50 ng/ml of rhBMP-2 (Fig. 4B). More than 10 μM of MEK-1 inhibitor significantly promoted ALP activity in MC3T3-E1 cells compared with that in the absence of MEK-1 inhibitor. The treatment with 25 μM of PD98059 increased ALP mRNA expression in the absence (data not shown) or presence of rhBMP-2 (Fig. 4D, left graph).
In contrast to the results using C2C12 cells, PD98059 enhanced OCN secretion from MC3T3-E1 cells in its dose-dependent fashion and peaked at 50 μM (0–50 μM; Fig. 4C). However, its secretion was reduced by the treatment with 100 μM of PD98059. This reduction might be caused by the cellular toxicity of this MEK-1 inhibitor. The increase of OCN mRNA expression was checked by RNA blot and OCN expression was consistent with its secretion between day 4 and 6 by the treatment with PD98059 (Fig. 4D, right graph and panels).
Calcium content of mineralized nodules was increased by the treatment with 25 μM of MEK-1 inhibitor. Twenty-five micromolars of PD98059 and 50 ng/ml of rhBMP-2 synergistically increased calcium content by 9.4-fold (Fig. 4 E).
Murine C2C12 cells represented myogenic differentiation under low-serum (5%) condition.(25) It also has been reported that the same cells were converted from myogenic differentiation into osteoblastic differentiation by stimulation with rhBMP-2 in the low-serum (5%) culture media(25,26) and adipogenesis by thiazolidinediones or fatty acids.(27,28) Here, we found osteoblastic differentiation of these cells by stimulation with rhBMP-2 even in the presence of 10% FBS, which is usually used for maintaining these cells as undifferentiation status. Therefore, we considered these cells as pluripotent mesenchymal stem cells and a model for osteoblastic differentiation by stimulation with BMP-2. MC3T3-E1 cells are thought to be committed to osteoblastic lineage. They were used as mesenchymal cells that were more differentiated into osteoblasts than C2C12 cells.
Then, we screened the small molecule compounds, especially kinase inhibitors, to stimulate osteogenesis in vitro. ALP activity of rhBMP-treated C2C12 cells and MC3T3-E1 cells was used as a marker for our screening. We picked up an MEK-1 inhibitor as a candidate for this condition. Another candidate we found was a Rho kinase inhibitor, Y-27632 (Yoshikawa H, Yoshioka K, Itoh K, submitted, 2002). In contrast, SB203580 p38 MAPK inhibitor did not promote and even inhibit the osteoblastic differentiation. In this report, we focused on the MEK-1 inhibitor PD98059.
Our present data summarized the following four results. First, MAPK signaling pathway was involved in the initiation of the differentiation of mesenchymal cells. MEK-1 inhibition stimulated myogenic differentiation of C2C12 cells even in the presence of 10% FBS when the cells were not committed to osteoblastic lineage by BMP-2, while it promoted the osteoblastic differentiation of BMP-2-stimulated C2C12 cell and MC3T3-E1 cells when committed to osteoblastic lineage. In addition, both myogenic and osteoblastic differentiation of C2C12 cells were shut down by the expression of continuous activation of MEK-1 (Fig. 3C; data not shown). Bennet et al.(29) reported that early myogenic differentiation was inhibited by overexpression of MAPK phosphatase 1 in C2C12 cells and they suggested that inactivation of MAPK might be required for C2C12 myoblasts to initiate myogenesis. Our results were consistent with their report and the inactivation of MAPK might be required for these cells to initiate the osteoblastic differentiation by BMP-2. MEK-1 inhibitor is thought to be an initiator or a promoter of the early stage differentiation of mesenchymal cells.
Second, investigating the osteoblastic differentiation of two cell lines in detail, ALP activity was increasing by the inhibition of MEK-1 activity in both C2C12 pluripotent mesenchymal cells treated with rhBMP-2 and MC3T3-E1 preosteoblastic cells with or without rhBMP-2 treatment. mRNA expression of type I collagen also was increasing by treatment with PD98059 (data not shown). Because ALP and type I collagen are used as early osteoblastic differentiation markers, these results indicated that MEK-1 inhibition promoted early osteoblastic differentiation in mesenchymal cells of both pluripotency and commitment into osteoblastic lineage. We speculate that two possibilities account for the reason for increase in ALP activity and type I collagen expression by MEK-1 inhibition. One idea is that the inhibition of MAPK signaling directly affects these expressions. The other is that MEK-1 inhibition promotes BMP production and subsequently induces these expressions. This issue should be focused on in future experiments.
Third, mineralization of ECM was promoted by the long-term inhibition of MAPK signaling pathway with MEK-1 inhibitor in both cell lines. It was reported that ALP cDNA-transfected cells promoted mineralized nodules.(30) Recently, Wennberg et al.(31) reported that osteoblasts derived from tissue nonspecific (TN) ALP knockout mice showed little mineralization of ECM and that mineralization was induced by adding soluble recombinant human TN ALP in these cells. They and other investigators suggested that ALP was involved in mineralization of ECM.(31,32) Taking our results and previous reports, the promotion of ALP activity might account for the effects of the MEK-1 inhibitor on the mineralization of ECM.
Fourth, OCN secretion and its transcriptional level are independent of MEK-1 inhibition and there is no correlation between its secretion and mineralization of ECM. OCN has been reported as a late osteoblastic differentiation marker.(2) However, bone formation was even increased in osteocalcin-deficient mice. In addition, bone mineralization and resorption were not impaired in those mice.(33) These reports suggested that OCN was a late osteoblastic marker but did not represent a degree of bone formation. In spite of the inconsistency of the effect of the MEK-1 inhibitor on OCN secretion and its transcript in between C2C12 and MC3T3-E1 cells, acceleration of mineralization of ECM in both cell lines is consistent with the results of OCN-deficient mice. Taking these facts, expression of OCN would not directly affect bone formation. Moreover, our data suggested that the difference between two cell lines caused the difference of OCN secretion by MEK inhibition. Xiao et al. reported that transcriptional factor Cbfa1, essential for the differentiation of osteoblasts, is phosphorylated by the activation of the MAPK pathway and regulated the activation of OCN gene promoter.(34) PD98059 did not affect Cbfa1 mRNA and protein expression level (data not shown). Further, Cbfa1 gene expression level was higher in MC3T3-E1 cells than in C2C12 cells, but OCN gene expression level was lower in the former than the latter (Figs. 2 and 4; data not shown). These data suggested that OCN gene expression was regulated not only by Cbfa1 but also by other signaling pathways affected directly or indirectly by MAPK signaling. Moreover, the same group recently reported that OCN gene expression was blocked by the short-term inhibition of the MAPK pathway by U0126 in confluent and differentiated MC3T3-E1 cells.(35) This report was partly inconsistent with our results. This difference might be due to the cellular background and/or the way of the application of the inhibitors.
Recently, Gallea et al. reported the positive effects of PD98059 on the osteogenic differentiation of C2C12 cells.(36) Our results were consistent with their report by means of ALP activity, OCN secretion, and these transcriptional levels. We found that the phosphorylation level of MAPK was reduced only slightly in the long-term inhibition of the MAPK pathway by high concentration of MEK-1 inhibitor because of the cellular compensatory mechanisms (Figs. 1A and 4A; data not shown). In addition to the experiments using the inhibitor, we could confirm those effects in C2C12 stable transfectants introducing constitutively active or dominant-negative MEK-1 cDNAs. Again, the change in MAPK phosphorylation was slight in these transfectants. Therefore, the cellular phenotype presented in this report represented the total (direct and indirect) effects of long-lasting inhibition of MAPK signaling. We also found that the MEK-1 inhibitor accelerated the mineralization of ECM, closely related to the bone formation in vivo. We also investigated that U0126(37) (Wako Pure Chemicals Industries, Ltd.) and U0124(37) (Calbiochem-Novabiochem Co., San Diego, CA, USA), another specific MEK-1 inhibitor and a negative control for U0126, showed similar effects on ALP activity and mineralization of ECM (data not shown). Our preliminary experiments, which were carried out to observe the effects of U0126 on in vivo ectopic bone formation by rhBMP-2 and type I collagen composite implanted submuscular fasciae in mice, showed enhanced ectopic bone formation to a certain degree (Higuchi C, Yoshioka K, Yoshikawa H, Itoh K, unpublished data, 2002).
Although the usefulness of the MEK-1 inhibitor for bone formation is clear, the precise underlying mechanism is not clear. The phosphorylation level of Smad5 or nuclear translocation of endogenous Smad5 was not influenced by the treatment with MEK-1 inhibitor (data not shown). Further studies should be required to solve the points.
This study suggested that a fine-tuning of the MAPK signaling pathway could promote mineralization of the ECM in parallel with up-regulation of ALP activity in BMP-2-induced pluripotent mesenchymal cells and preosteoblastic cells in vitro. MEK-1 inhibitors would be useful in vivo for the promotion of bone formation, for instance, delayed fracture healing or focal osteoporotic change, around fracture site.
This study was supported in part by grants-in-aid from the Ministry of Education, Science, Sports, and Culture of Japan and by an industrial technology research grant in 2000 from the New Energy and Industrial Technology Development Organization of Japan as well as by research grants from the Yamanouchi Foundation for Research on Metabolic Disease and the Takeda Science Foundation.